Investigating Change in Intraindividual Factor Structure Over Time

Although changes in factor structure are sometimes seen as hindrances to scientific research, the investigation of factor structure changes can also provide new opportunities for the development of theory in psychology. Some researchers (e.g., Coan, 1966; Emmerich, 1968; Baltes & Nesselroade, 1973) have noted the importance of investigating structural change within psychology when studying such topics as factor integration, differentiation, emergence, and disintegration. The methods proposed to investigate such structural changes over time are extensions of P-technique factor analysis (Cattell, 1947, 1952) used to examine change in the factor model parameters over time at the individual level. In particular, ``local' P-technique factor models are fit within relatively small time windows followed by growth modeling of the factor model parameter trajectories. A Monte Carlo simulation study was undertaken to evaluate the estimation and model evaluation procedures proposed for investigating changes in intraindividual factor structure over time. The meaning and implications of these initial results are discussed with the hope of stirring thought on the potential scientific impact of examining changes in intraindividual factor structure over time along with future development of parameter estimation and model evaluation procedures.